Generated banner for Matlab Process Images
Generated logo for Matlab Process Images

Matlab Process Images

MathWorksOfficialCreated by MathWorks

  • Skill
  • Tool integration
  • Software development
  • BSD-2-Clause
  • No issues found

From the author

Load this first for any MATLAB task involving images, photos, scans, frames, volumes, or visual data — reading, writing, filtering, enhancing, denoising, sharpening, deblurring, thresholding, segmenting, masking, detecting, counting objects, removing backgrounds, measuring, registering, stitching, aligning, cropping, resizing, or transforming. Covers image I/O (TIFF, PNG, JPEG, RAW, HDR), histograms, edge detection, color, morphology, deep learning for images, quality assessment, and 3-D volumes. Also for image problems (dark, noisy, blurry, low contrast, uneven lighting, hazy, color cast) or data exploration (folder of images). Includes medical imaging I/O (DICOM, NIfTI, NRRD, medicalVolume) and large/blocked images (blockedImage, whole-slide). Skip if the user just wants you to look at an image, with no MATLAB work. Not for Simulink video-block models, C/C++/CUDA code generation or GPU deployment of a network, or point clouds. Hyperspectral and visual inspection have sibling skills.
— the description by MathWorks

AutoGineer summary

Written by AutoGineer from the skill’s own files, not by the author

This skill serves as the primary entry point for MATLAB tasks involving images, scans, frames, and 3-D volumes. It provides core conventions for image data types, coordinate systems, and display practices, while routing users to specific references for tasks such as I/O, filtering, segmentation, and measurement. It ensures consistent use of Image Processing Toolbox functions and helps users navigate common pitfalls in image analysis workflows.

What it does

  • Provides standardized conventions for image data types and coordinate systems.
  • Routes users to specialized references for image I/O, filtering, and transformation tasks.
  • Defines best practices for displaying images using the imageshow function.
  • Offers guidance on handling 3-D volumetric data and large-scale images.
  • Assists with selecting appropriate functions for segmentation and object measurement.
  • Provides troubleshooting patterns for common image processing errors.

When to use it

  • When performing image processing tasks in MATLAB, such as filtering, enhancing, or segmenting.
  • When working with image formats like TIFF, PNG, JPEG, DICOM, or NIfTI.
  • When needing to understand MATLAB-specific coordinate conventions for image data.
  • When determining the correct toolbox function for a specific image analysis goal.
  • When troubleshooting syntax or type-related errors in image processing scripts.

Advantages

  • Ensures consistent coding patterns across different image processing tasks.
  • Reduces errors by providing clear guidance on coordinate and data type conventions.
  • Simplifies the selection of appropriate tools for complex workflows like 3-D volume processing.
  • Promotes the use of modern display functions like imageshow over legacy alternatives.
  • Provides a structured approach to identifying the correct documentation for specific functions.

Usage examples

“How do I read a DICOM file and display it in MATLAB?”
The agent will guide you to use the medical-data-io reference for reading the file and the imageshow function for display, following the core conventions.
“I need to count the number of objects in a binary image.”
The agent will provide the standard analysis pipeline using bwconncomp and regionprops, ensuring the input is correctly formatted as logical.
“How do I perform 3-D filtering on a volume?”
The agent will direct you to the 3-D volume processing reference to select the appropriate 3-D specific function, such as imgaussfilt3.
“Why is my regionprops measurement returning only one object?”
The agent will explain that the input must be a logical array and provide the code to convert your numeric array to logical.

Install

  1. Download the skill as a ZIP:

    https://autogineer.com/agent-skills/matlab/matlab-process-images/download
  2. Unzip it into the cross-client skills folder of your project:

    mkdir -p .agents/skills && unzip matlab-process-images.zip -d .agents/skills

Documentation by MathWorks

Image Processing

Process, analyze, and transform images using MATLAB Image Processing Toolbox (IPT). This skill routes you to the correct reference for your task and provides core conventions that apply across all IPT work.

When to Use

Loaded for general-purpose IPT work: file and batch I/O, filtering, restoration, morphology, segmentation, measurement, geometric transforms, color, texture, quality metrics, deep learning on images, and image apps. Non-medical microscopy — cell and particle counting on plain IPT — is in scope.

Hand off (and co-load) instead of answering alone: see the handoff table under ## Routing.

When NOT to Use

  • Training a network with no image-specific preprocessing — use Deep Learning Toolbox directly

  • Simulink or Stateflow video blocks — use Simulink directly

  • C/C++ code generation, MEX, GPU acceleration and deployment — use MATLAB Coder / GPU Coder directly

  • Point clouds, 3-D reconstruction, and video stabilization — use Computer Vision Toolbox directly

  • General MATLAB questions unrelated to images

  • Looking at an image yourself (e.g., "look at this screenshot") when no MATLAB work is requested — use your own vision instead

Core Conventions

Image Data Types
TypeRangeWhen to use
uint80–255Display, file I/O, memory-efficient storage
uint160–65535High dynamic range sensors, medical. Camera RAW is n-bit in a uint16 container — rescale with rawinfo BlackLevel/WhiteLevel, not 65535
single / double0.0–1.0Arithmetic, filtering, computation
logical0 or 1Binary masks, segmentation results

Most IPT functions accept integer types directly — do not convert to floating-point unless required.

Default Coordinate Conventions
  • MATLAB images are (row, col) — row is Y (down), column is X (right)

  • size(img) returns [height, width, channels]

  • Spatial functions use (x, y) order for coordinates (e.g., imref2d, regionprops centroids)

  • Pixel centers are at integer intrinsic coordinates; in default world referencing, pixel (1,1) center is at world (1.0, 1.0) and the image edge starts at 0.5

  • imref2d and imref3d can specify non-default spatial referencing to place images into world coordinate systems. By default they use the unit-spaced, integer-pixel-centered coordinate system described here, e.g. Rout = imref2d(imageSize). Functions that do not accept spatial referencing objects assume this intrinsic coordinate system for any operation requiring spatial referencing.

Multichannel Images
  • Grayscale: M×N or M×N×1

  • RGB: M×N×3

  • With alpha: read using [img, ~, alpha] = imread(...) (alpha is the third output)

  • Convert: rgb2gray, im2gray (preferred, handles more input types)

Display Functions — use imageshow, not imshow

Always use imageshow instead of imshow in all generated code — scripts, apps, examples, and quick previews. The signatures are compatible for common cases: replace imshow(...) with imageshow(...). imageshow (R2024b+) handles sub-bit-depth containers (e.g., 12-bit in uint16) via DisplayRangeMode, supports OverlayData for label/mask overlays, and works with viewer2d. imshow lacks all of these capabilities and should not appear in new code.

Title convention: imageshow creates a viewer2d, not an axes — gca does not return it. To add a title, get the viewer from the Image object's parent:

obj = imageshow(im);
viewer = obj.Parent;
title(viewer, "Title Text");

Do not use title("text") or title(gca, "text") after imageshow — these silently fail or create a stray axes title.

Code Execution

Execute code blocks via evaluate_matlab_code before presenting them to the user. If it errors, fix and re-run. This catches type mismatches, wrong argument orders, and nonexistent functions before the user sees them. If execution requires user data or toolboxes not available in the current session, present the code with a note that it should be run with the user's own data — do not loop trying to synthesize stand-in data.

Error Recovery

If code throws an error after execution, load the matlab-read-documentation skill and look up the function's doc page to verify syntax, required arguments, and supported input types before retrying.

When to Ask the User

Ask for clarification before proceeding when:

  • The image type/format is ambiguous and affects the processing approach

  • Multiple valid approaches exist with different tradeoffs (speed vs quality, lossy vs lossless)

  • The task requires choosing parameters that depend on domain knowledge (e.g., object size thresholds, noise characteristics)

  • The user's MATLAB release is unknown and the recommended function may not be available

Routing

Match your task in the tables below. Check the handoff table first. If the task belongs to another skill, load that skill in addition to this one — the Core Conventions above (data types, coordinate order, code execution) apply to all image data. Do not re-derive them from the sibling. If a row in both tables fits, the handoff table wins on workflow; this skill still owns conventions.

  • Unfamiliar data? Load references/image-understanding.md first — but only when class, range, or content is unknown.

  • Starting from a function name, not a task? The function column is illustrative, not exhaustive. To locate any function: grep -rln '<functionName>' references/. To look up syntax details, load matlab-read-documentation.

  • Pipeline tasks: load several references; cap at three unless the task genuinely spans more.

  • Follow the patterns in loaded references. References contain specific code patterns and workflows — use them instead of improvising equivalent code from general knowledge.

  • No row matches? Name the owning product, state the closest IPT fallback, and say so — do not guess a reference.

Check first: is this another skill's job?
If the task involves...Also load
Segmenting anatomy — tumor, lesion, organ, vessel — in CT, MRI, PET, ultrasound; MedSAM; radiomicsNo skill covers this — Medical Imaging Toolbox (segmentMedicalImage, groundTruthMedical); look up the doc page
Displaying or viewing a 2-D image: imshow, imageshow, viewer2d, montage, mask overlay, annotation, colorbar, colormap, figure exportmatlab-display-image
Displaying a 3-D volume: volshow, viewer3d, isosurface rendering, slice viewer, orthogonal viewsmatlab-display-volume
Hyperspectral or multispectral cubes, spectral indices, unmixing, ENVI, band selectionmatlab-analyze-spectral-images
Calling a PyTorch or ONNX vision model from MATLABmatlab-integrate-pytorch-vision
OCR, text recognition, reading text out of an imagematlab-recognize-text (requires Computer Vision Toolbox)
Factory-floor visual inspection: anomaly detection, defect/scratch/dent detection, gauging, pass/fail countingmatlab-use-visual-inspection (requires Visual Inspection Toolbox, R2026b+)
Barcode or QR reading, camera calibrationNo skill covers this — Computer Vision Toolbox (readBarcode, estimateCameraParameters); look up the doc page
Optical system design: lens import, prescription-side MTF, spot diagrams, ray tracing, tolerance analysismatlab-model-optics
In scope: pick the reference
If the task or the user's words involve...Illustrative functionsLoad
inspect, understand, characterize, explore data, look at my images, what kind of image, what am I working with, what's in this folder, describe the data, summarize the dataset, what format, bit depth, class and range, indexed image, alpha channel, metadata, file info, dataset characterization, before I start processingimfinfo, imhist, countEachLabelreferences/image-understanding.md
read, write, load, save, import, export, open an image, convert format, batch, folder of images, TIFF, PNG, JPEG, RAW, HDR, test patternimread, imwrite, imageDatastore, hdrreadreferences/image-io.md
video, footage, frames, read video, write video, VideoReader, VideoWriter, frame rate, video codec, MP4, AVIVideoReader, VideoWriter, readFrame, hasFrame, writeVideoreferences/video-io.md
smooth, blur an image, sharpen, denoise, noise, grainy, filter, kernel, padding, edge-preserving, bilateral, non-local means, frequency filterimgaussfilt, medfilt2, imsharpen, imbilatfilt, imnlmfilt, imdiffusefilt, imfilter, ordfilt2, locallapfilt, wiener2references/filtering.md
contrast, brighten, darken, too dark, washed out, enhance, improve, histogram, equalize, CLAHE, normalize, rescale, blend, arithmetic, type convert, flat-field, illumination correction, vignetting, uneven lightingimadjust, stretchlim, adapthisteq, histeq, rescale, mat2gray, im2single, imadd, imlincomb, imblend, imflatfieldreferences/preprocessing.md
erode, dilate, open, close, structuring element, 2-D morphology, fill holes, remove small objects, skeletonize, thin, tidy up a binary resultstrel, imerode, imdilate, imopen, imclose, imfill, bwareaopen, bwskel, imtophat, imreconstructreferences/morphological-ops.md
deblur, blurry, unfocused, out of focus, motion blur, camera shake, restore a blurred image, deconvolution, PSF estimationdeconvblind, deconvlucy, deconvwnr, deconvregreferences/deblurring.md
resize, rotate, crop, warp, scale, flip, straighten, affine, projective, perspective, undistort, lens distortion, spatial referencing, world coordinates, voxel coordinates, pixel coordinates, coordinate conversionimresize, imrotate, imcrop, imwarp, imtranslate, affinetform2d, imref2dreferences/geometric-transforms.md
align two images, stitch, mosaic, panorama, overlapping shots, seam, motion correction, multi-modal alignment, control points, non-rigid, deformable, homography (feature-based requires Computer Vision Toolbox)imregister, imregtform, imregdemons, imregcorr, detectSIFTFeatures, matchFeatures, estgeotform2d (last three: Computer Vision Toolbox), imregdeform (Medical Imaging Toolbox)references/registration.md
segment or threshold with no user input: binarize, Otsu, watershed, k-means or colour clustering, imsegsam automatic mode, deep-learning segmentationimbinarize, graythresh, multithresh, adaptthresh, watershed, imsegkmeans, imsegkmeans3, imsegisodata, imsegsamreferences/segmentation-automated.md
segment with user input: clicks, points, boxes, scribbles, seeds, prompted SAM, grabcut, active contours, region growing, lazy snapping, superpixels, App Designer segmentation UIssegmentAnythingModel, grabcut, activecontour, lazysnapping, imsegfmm, grayconnected, superpixelsreferences/segmentation-interactive.md
measure, count objects, how many, area, perimeter, centroid, size, shape, width, thickness, gap, diameter, Feret, orientation, detect edges, circles, lines, boundary tracing, ROI, world units, metrology, region, region properties, analysisregionprops, bwconncomp, bwareafilt, bwpropfilt, bwlabel, edge, imfindcircles, hough, bwboundariesreferences/analysis-measurement.md
quality, PSNR, SSIM, sharpness, noise level, is it blurry, compare two images, difference, before/after, MTF (camera/chart SFR), test chart, artifact (register first — full-reference metrics need pixel correspondence)psnr, ssim, multissim, immse, brisque, niqe, piqe, measureSharpness, esfrChartreferences/quality-assessment.md
texture, repetitive pattern, frequency domain, Fourier, FFT, DCT, Radon, Gabor, co-occurrence, GLCMgraycomatrix, graycoprops, entropyfilt, fft2, dct2, radon, gaborreferences/texture-transforms.md
color space, LAB, HSV, hue, saturation, white balance, color cast, color correction, color balance, ICC profile, grayscale conversionrgb2lab, lab2rgb, rgb2hsv, chromadapt, illumwhite, makecform, deltaE, decorrstretchreferences/color-processing.md
3-D array, volume, volumetric, voxel, slice stack, z-stack — filtering, 3-D morphology, 3-D transforms, or measurement applied to a volumeimgaussfilt3, medfilt3, imresize3, regionprops3, bwvolumefilt (R2026b), bwpropfilt3 (R2026b)references/3d-volume-processing.md
deep learning, neural network, train, U-Net, semantic segmentation networks, classify images, pretrained model, transfer learning, fine-tune, inference, augment, training dataset (requires DLT, often CVT)unet, deeplabv3plus, trainnet, semanticseg, imageDataAugmenter, augmentedImageDatastorereferences/deep-learning-imaging.md
build your own image app or GUI: sliders, live preview, parameter tuning, ROI widgetsuifigure, uigridlayout, uislider, drawrectanglereferences/app-building.md
DICOM, NIfTI, NRRD, DICOM-RT, patient metadata, anonymizing, medicalVolume, medicalImage, reading/writing medical formats, extractSlice, volume orientation, medical spatial referencingmedicalVolume, medicalImage, dicomreadVolume, dicomCollection, extractSlice, dicomContours, dicomanonreferences/medical-data-io.md
gigapixel, whole-slide, out-of-memory images, blockedImage, tiled workflows, image pyramids, .svs, .ndpi, .czi, TIFF > 10,000 px, parallel image processing, custom adaptersblockedImage, apply, getBlock, getRegion, selectBlockLocations, blockedImageDatastorereferences/large-image-processing.md

Toolbox Detection

When unsure what's available, check installed toolboxes:

hasIPT = ~isempty(ver('images'));
hasCVT = ~isempty(ver('vision'));
hasDLT = ~isempty(ver('nnet'));
hasMIT = ~isempty(ver('medical'));
hasWavelet = ~isempty(ver('wavelet'));

Most functions in this skill require Image Processing Toolbox. The deep-learning-imaging reference also requires Deep Learning Toolbox. Feature-based registration (detectSIFTFeatures, matchFeatures, estgeotform2d) requires Computer Vision Toolbox.

Release-Gated Features

This skill targets R2024b+ but some features require newer releases. References annotate these with the required release. Key ones:

  • R2024b: imageshow, imblend, viewer2d

  • R2026a: uidraw, uipaint, linkviewers

  • R2026b: bwvolumefilt, bwpropfilt3, uiannotate

If targeting an older release, check with isMATLABReleaseOlderThan("R2026a") and use the fallback patterns documented in the references. To check release notes for a specific function, load matlab-read-documentation.


Copyright 2026 The MathWorks, Inc.


Files

references/segmentation-interactive.md

Interactive Segmentation

Decision guidance for segmentation methods that use user input — point/box prompts, scribbles, ROIs, seed points, or initial masks. Prompted SAM is a strong general-purpose option for point or bounding-box prompts; classical methods may be more appropriate when the task is defined by intensity similarity, edge contrast, or an initial mask.

Often loaded with: morphological-ops.md (cleaning up masks), analysis-measurement.md (measuring segmented regions), segmentation-automated.md (when no user input is available).

Method Selection

TaskUser inputAlgorithm
Segment an object from sparse visual promptsPoint or bounding boxsegmentAnythingModel (prompted SAM 2 / SAM) (CVT+DLT)
Foreground/background separationROI rectanglegrabcut (requires superpixels)
Foreground/background from strokesScribbleslazysnapping (requires superpixels)
2–3 regions from strokesScribblesimseggeodesic
Flood-fill from seedSeed point + tolerancegrayconnected
Region growing from seedsSeed pointsimsegfmm
Evolving boundary to edgesInitial maskactivecontour
Medical organ/tumor (interactive box)Bounding boxmedicalSegmentAnythingModel (MIT)
Cell/nucleus instancesAutomatic or promptedcellpose → segmentCells2D / segmentCells3D (MIT)

General-purpose recommendation: Use prompted SAM when available — it handles a wide range of scenarios with minimal user input and no training data. For tasks defined by intensity similarity (flood-fill, region growing) or evolving an initial mask to edges, the classical methods below may be more direct.

Coordinate order varies by function: SAM functions (ForegroundPoints, BoundingBox) use (x, y) order. grayconnected uses (row, col) order. graydiffweight and coordinate-based imsegfmm use (col, row) order (column first, then row). Active contour and masks use standard (row, col) array indexing.

Prompted SAM 2 and SAM (R2024b+)

Requires IPT Model for Segment Anything Model support package and Computer Vision Toolbox + Deep Learning Toolbox. By default, segmentAnythingModel uses "sam2-large" but also supports other SAM 2 variants (e.g., "sam2-tiny", "sam2-small", "sam2-base-plus") and the base SAM model ("sam-base").

try
    sam = segmentAnythingModel;  % Verifies support package and required add-on are available
    hasSAM = true;
catch
    hasSAM = false;  % Help the user install the support package
end

Do not use which alone — it can return true even when required add-ons are missing. Attempt construction to confirm usability.

Setup
sam = segmentAnythingModel;              % Uses "sam2-large" by default
embeddings = extractEmbeddings(sam, I);  % Expensive — do ONCE per image
Point Prompts
% Single foreground point — [x, y] coordinates on the object
masks = segmentObjectsFromEmbeddings(sam, embeddings, size(I), ...
    ForegroundPoints=[200 150]);

% Refine with background points
masks = segmentObjectsFromEmbeddings(sam, embeddings, size(I), ...
    ForegroundPoints=[200 150], BackgroundPoints=[50 50; 350 350]);
Bounding Box Prompt
% [x, y, width, height]
masks = segmentObjectsFromEmbeddings(sam, embeddings, size(I), ...
    BoundingBox=[50 80 200 180]);
Multiple Objects from Same Image
% Extract embeddings once, segment multiple times
embeddings = extractEmbeddings(sam, I);

mask1 = segmentObjectsFromEmbeddings(sam, embeddings, size(I), ...
    ForegroundPoints=[100 200]);
mask2 = segmentObjectsFromEmbeddings(sam, embeddings, size(I), ...
    BoundingBox=[300 100 150 200]);

Key rules:

  • segmentObjectsFromEmbeddings returns an H×W logical mask (single best mask by default). Use ReturnMultiMask=true to get H×W×3 with three ranked candidates.

  • Extract embeddings once per image, segment multiple objects from same embeddings

  • size(I) (third argument) is required — maps embeddings back to image coordinates

  • ForegroundPoints and BoundingBox use (x, y) order, not (row, col)

  • Combine foreground + background points for difficult cases

When SAM is unavailable: Use grabcut for foreground/background, activecontour for boundary refinement, or imsegfmm for region growing.

GrabCut — Foreground/Background from ROI

Gotcha: grabcut requires a superpixel label matrix in addition to the image. Always compute superpixels first:

[L, N] = superpixels(I, 500);    % REQUIRED before grabcut
ROI = false(size(I,1), size(I,2));
ROI(50:300, 100:350) = true;
BW = grabcut(I, L, ROI);

Works on 3-D volumes too (with superpixels3).

Lazy Snapping — Foreground/Background from Scribbles

Requires scribble masks indicating foreground and background regions. fgScribble and bgScribble are logical masks (same size as image) where true marks user-drawn strokes:

[L, N] = superpixels(I, 500);    % REQUIRED before lazysnapping
BW = lazysnapping(I, L, fgScribble, bgScribble);

Geodesic Segmentation — 2–3 Regions from Scribbles

For partitioning an image into a small number of regions based on user-drawn scribbles. Requires RGB input — does not work on grayscale. Returns a label matrix (not a binary mask):

L = imseggeodesic(RGB, scribble1, scribble2);           % 2 regions
L = imseggeodesic(RGB, scribble1, scribble2, scribble3); % 3 regions

Gray Connected — Flood-Fill from Seed

Simple region growing based on intensity similarity to a seed pixel:

BW = grayconnected(I, row, col, 20);  % tolerance is positional (not a Name-Value pair)

% R2026b: restrict to 4-connected pixels (horizontal/vertical only)
BW = grayconnected(I, row, col, 20, Connectivity=4);

Active Contours — Evolving Boundary

Evolves an initial mask toward object boundaries. Users struggle most with creating the initial mask. Options:

% From rough threshold + dilate (most common)
mask = imbinarize(I);
mask = imdilate(mask, strel("disk", 10));
BW = activecontour(I, mask, 200, "Chan-Vese");

% From bounding rectangle
mask = false(size(I));
mask(r1:r2, c1:c2) = true;
BW = activecontour(I, mask, 200, "edge");
MethodUse when
"Chan-Vese"Region-based, objects without strong edges
"edge"Objects with clear boundary edges

Works on 3-D volumes with same syntax.

Fast Marching — imsegfmm

Region growing from seeds using a weight array:

% Weight from intensity difference to seed
% graydiffweight and imsegfmm take (C, R) order — column first, then row
W = graydiffweight(I, col, row, GrayDifferenceCutoff=25);
BW = imsegfmm(W, col, row, 0.01);  % threshold on travel time

% Weight from gradient magnitude
W = gradientweight(I);
BW = imsegfmm(W, seedMask, 0.01);

Medical Segmentation (Medical Imaging Toolbox)

For medical images with MIT available:

TaskFunction
2-D organ/tumor (interactive box)medicalSegmentAnythingModel
3-D automatic (organs)medicalImageLabeler with MONAI Label
Cell/nucleus segmentationcellpose → segmentCells2D / segmentCells3D
Radiomic features from ROIradiomics

For full medical segmentation workflows (MedSAM, Cellpose, radiomics), see Medical Imaging Toolbox documentation.

Building Interactive Segmentation Apps

The Image Segmenter app (imageSegmenter) provides built-in interactive segmentation tools, including SAM. Use it for ad-hoc interactive segmentation sessions.

For custom apps, use App Designer with SAM. The pattern combines a viewer2d for display, annotation-based prompts, and real-time mask preview. For a complete walkthrough, see the doc example "Create an App to Segment Images with SAM" (R2026b).

Architecture Pattern

Build a custom UI component inheriting from matlab.ui.componentcontainer.ComponentContainer:

classdef SAMSegmenter < matlab.ui.componentcontainer.ComponentContainer
    properties
        Image
        Mask
        SavedMask
        SegmentationAlpha = 0.5
    end
    properties (Access = private)
        Model           % segmentAnythingModel object
        Embeddings      % Extracted once per image
        Viewer          % images.ui.graphics.Viewer
        ImageObject     % Image displayed in viewer
        Bbox            % Current bounding box
        SegmentRequired % Flag to trigger update
        CurrentLabelID = 1
        IsForegroundMode = true
    end
    methods (Access = protected)
        function setup(comp)
            comp.Viewer = viewer2d(comp);
            comp.ImageObject = imageshow([], Parent=comp.Viewer);
            comp.Viewer.AnnotationAddedFcn = @(~,evt) annotationAdded(comp, evt);
        end
    end
end
Key Workflow: Load → Embed → Annotate → Preview → Commit
% 1. Load image and extract embeddings ONCE
function loadImage(comp, imageData)
    comp.Image = imageData;
    comp.Model = segmentAnythingModel("sam2-large");
    comp.Embeddings = extractEmbeddings(comp.Model, imageData);
    comp.ImageObject.Data = imageData;
    comp.SavedMask = zeros(size(imageData, [1 2]), "uint8");
end
Annotation-Driven Segmentation (update method)

The update method fires whenever annotations change, providing real-time preview:

function update(comp)
    if ~comp.SegmentRequired, return; end
    comp.SegmentRequired = false;

    imageSize = size(comp.ImageObject.Data);
    annotations = comp.Viewer.Annotations;

    % Collect foreground/background points from annotations
    foreground = [];
    background = [];
    for idx = 1:numel(annotations)
        if isa(annotations(idx), "images.ui.graphics.roi.Point") && isvalid(annotations(idx))
            if annotations(idx).UserData  % true = foreground
                foreground(end+1,:) = annotations(idx).Position(1:2);
            else
                background(end+1,:) = annotations(idx).Position(1:2);
            end
        end
    end

    if isempty(foreground) && isempty(comp.Bbox), return; end

    % Segment using current prompts
    mask = segmentObjectsFromEmbeddings(comp.Model, comp.Embeddings, imageSize, ...
        ForegroundPoints=foreground, BoundingBox=comp.Bbox, BackgroundPoints=background);

    % Preview: overlay new mask on saved mask
    priorMask = comp.SavedMask;
    priorMask(mask) = uint8(comp.CurrentLabelID);
    comp.ImageObject.OverlayData = priorMask;
end
Handling Annotations via Viewer Events

Use the AnnotationAdded event from Viewer to process user interactions:

function annotationAdded(comp, event)
    event.Annotation.Interactions = "click";

    if isa(event.Annotation, "images.ui.graphics.roi.Rectangle")
        % Bounding box drawn — save position, switch to point mode
        comp.Bbox = event.Annotation.Position;
        comp.Viewer.Mode.Annotate.Style = "point";
        comp.Viewer.Mode.Annotate.ContinueAnnotating = true;
        comp.Viewer.Mode.CurrentMode = "annotate";
    else
        % Point — validate it's inside bounding box
        if ~isempty(comp.Bbox)
            pos = event.Annotation.Position;
            if pos(1) < comp.Bbox(1) || pos(2) < comp.Bbox(2) || ...
                    pos(1) > (comp.Bbox(1)+comp.Bbox(3)) || pos(2) > (comp.Bbox(2)+comp.Bbox(4))
                delete(event.Annotation);
                return;
            end
        end
        event.Annotation.UserData = comp.IsForegroundMode;
    end
    comp.SegmentRequired = true;
end
Commit / Revert Pattern
% Commit current preview to saved mask
function saveMask(comp)
    comp.SavedMask = comp.ImageObject.OverlayData;
    comp.Mask = comp.SavedMask;
    comp.Viewer.Annotations = [];
    comp.Bbox = [];
end

% Revert to last saved state
function clearAll(comp)
    comp.Viewer.Annotations = [];
    comp.ImageObject.OverlayData = comp.SavedMask;
    comp.Bbox = [];
end
App-Level Integration

Wrap the component in an App Designer app with toolstrip controls:

function startupFcn(app)
    app.SamSegmenter = SAMSegmenter(app.WorkingAreaGrid);
end

function loadImageFromFile(app)
    [file, location] = uigetfile("*.*", "Select an image");
    if ~isequal(file, 0)
        d = uiprogressdlg(app.UIFigure, Title="Please Wait", ...
            Message="Generating SAM embeddings...", Indeterminate="on");
        imageData = imread(fullfile(location, file));
        loadImage(app.SamSegmenter, imageData);
        close(d);
    end
end

function exportMask(app)
    mask = app.SamSegmenter.Mask;
    assignin("base", "SegmentedImage", mask);
end
Design Guidelines
  • Extract embeddings once per image — show a progress dialog since it's expensive

  • Use Viewer with OverlayData for real-time mask overlay (not labeloverlay + redraw)

  • Validate point positions — discard points outside the active bounding box

  • Commit/revert pattern — let users build up multi-label masks incrementally

  • ComponentContainer inheritance — isolates segmentation logic from app layout for reusability

  • Use Viewer.Mode.Annotate to control annotation style (point vs rectangle) and continuation

This pattern extends beyond SAM — replace the segmentObjectsFromEmbeddings call with any interactive segmentation method (grabcut, activecontour, imsegfmm, etc.) to build custom annotation apps for any interactive segmentation workflow.

Displaying Segmentation Results

% Direct overlay display (preferred — one step, interactive)
imageshow(I, OverlayData=BW);

% For generating an RGB array (e.g., saving to file or further processing)
rgb = labeloverlay(I, BW, Transparency=0.5);

Conventions

  • Always compute superpixels before grabcut or lazysnapping

  • For prompted SAM, extract embeddings once per image — never per query

  • Use bwareafilt or imfill to clean up after interactive segmentation (see morphological-ops.md)

  • Prompted SAM is a strong default when CVT+DLT are available; use classical methods when the task is better defined by intensity, edges, or an initial mask

Common Mistakes

MistakeWhy it's wrongCorrect approach
grabcut(I, [], ROI) without superpixelsRequires label matrix input[L,N] = superpixels(I,500) then grabcut(I,L,ROI)
extractEmbeddings per bounding boxEmbeddings are expensiveExtract once, call segmentObjectsFromEmbeddings multiple times
activecontour without initial maskRequires a mask to evolve fromThreshold + dilate, or bounding rectangle
regiongrowing(I, seed)Does not exist in IPTUse activecontour, imsegfmm, or grayconnected
SAM coordinates in (row, col)SAM uses (x, y) orderForegroundPoints=[x y], not [row col]
graydiffweight(I, row, col) or imsegfmm(W, row, col, t)These take (C, R) order — column first, then rowgraydiffweight(I, col, row), imsegfmm(W, col, row, t)
grayconnected(I, r, c, Tolerance=20)Tolerance is positional, not a Name-Value pairgrayconnected(I, r, c, 20)
imseggeodesic(grayImg, ...)Requires RGB input; returns label matrix, not binaryL = imseggeodesic(RGB, scribble1, scribble2)
imshow(I, OverlayData=BW)imshow does not support OverlayDataimageshow(I, OverlayData=BW) (R2024b+)

Copyright 2026 The MathWorks, Inc.


Requirements and compatibility

Works with: Claude Code, Claude apps, OpenAI Codex, GitHub Copilot, Cursor, Gemini CLI, Kiro, OpenCode, Goose, Amp, Roo Code, JetBrains Junie, Factory Droid, Devin (ex-Windsurf), Mistral Vibe, OpenHands, Letta

“Works with” means the tool reads the open Agent Skills format; a skill written for one tool may use features others lack.

Security

No issues found by our static scan· No issues found

A static scan is not a guarantee. Skills are third-party content: review a skill before you let an agent use it.

Attribution and licence

Created by
MathWorks
Publisher
MathWorks (Official)
Source
matlab/matlab-agentic-toolkit · skills-catalog/image-processing-and-computer-vision/matlab-process-images · commit 2a7f215
Licence
BSD 2-Clause License (redistribution allowed)
  • LICENSE (added by AutoGineer from LICENSE.md)
  • ATTRIBUTION.md (written by AutoGineer)
Synced

AutoGineer is not affiliated with or endorsed by MathWorks or MathWorks.

Download ZIP

The ZIP holds the skill’s files, its licence files and an ATTRIBUTION.md. Download matlab-process-images.zip